← SIG (Susquehanna) Interview Insights

SIG (Susquehanna)·Data Scientist·Technical Phone Screen·Senior

Senior
May 2026

Summary

SIG had me work through a pretty dense stats and experimentation problem set built around a delivery platform scenario. Five sub-questions covering sample size math, non-inferiority testing, multiple comparison corrections, cluster randomization, and sequential monitoring. No metadata on outcome or timing was shared.

Questions Asked (5)

Q1

Given a baseline mean delivery time of 42 minutes with SD of 15 minutes and a target relative improvement of negative 5%, compute the per-arm sample size for a two-sample t-test using alpha 0.05 two-sided and 80% power.

A/B Testing & ExperimentationProduct Analytics & Metrics
Author's notes

The mechanics here are standard: delta is 0.05 * 42 = 2.1 minutes, then plug into n = 2 * ((z_alpha/2 + z_beta) / delta * sigma)^2.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

First, translate the relative improvement into an absolute effect size by multiplying the baseline mean by 5% (since negative improvement means a reduction). Then, use the standard formula for sample size per arm in a two-sample t-test: n = 2*(Z_{α/2} + Z_β)^2 * σ^2 / Δ^2, where Z_{α/2}=1.96 and Z_β=0.84 for 80% power. Finally, compute the numerical value and round up to the nearest integer.

Pro tip: Always clarify whether the relative improvement is a reduction or increase, and explicitly state that you're assuming equal variances and a two-sided test. Also, mention that this is a simplified calculation and that in practice you might use simulation or adjust for unequal variances.

1. Interpret the relative improvement

Calculate the absolute change: 5% of 42 minutes = 2.1 minutes. Since it's a negative improvement (reduction), the target mean is 42 - 2.1 = 39.9 minutes, so the effect size Δ = 2.1 minutes.

2. Identify the formula and parameters

For a two-sample t-test with equal variances, the per-arm sample size formula is n = 2*(Z_{α/2} + Z_β)^2 * σ^2 / Δ^2. Here, α=0.05 two-sided so Z_{α/2}=1.96, power=80% so Z_β=0.84, σ=15, and Δ=2.1.

3. Plug in values and compute

Compute (1.96+0.84)^2 = 2.8^2 = 7.84. Then n = 2 * 7.84 * (15^2) / (2.1^2) = 2 * 7.84 * 225 / 4.41. Calculate: 2*7.84=15.68; 15.68*225=3528; 3528/4.41 ≈ 800. So n ≈ 800 per arm.

4. Round up and state assumptions

Round up to the next integer: 800 per arm. Mention assumptions: equal variances, normal distribution, and that this is a simplified calculation (e.g., ignoring continuity correction or unequal allocation).

Key Points to Mention

  • Relative vs absolute effect size: 5% of 42 minutes is 2.1 minutes.
  • Two-sample t-test sample size formula: n = 2*(Z_{α/2}+Z_β)^2 * σ^2 / Δ^2.
  • Z-values: 1.96 for 95% confidence (two-sided), 0.84 for 80% power.
  • Assumptions: equal variances, normality, independent samples.
  • Rounding up to ensure sufficient power.
  • Practical considerations: potential need for larger sample due to non-compliance or missing data.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q2

For a non-inferiority test on cancellation rates where the baseline is 6% and the acceptable margin is 0.5 percentage points, compute the per-arm sample size using one-sided alpha of 0.05 and 80% power.

A/B Testing & ExperimentationProduct Analytics & Metrics
Author's notes

Non-inferiority framing always trips me up a little because the null and alternative are flipped from what you'd normally write.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

First, clarify that this is a non-inferiority test comparing a new treatment to a control with a margin of 0.5 percentage points, and that the baseline cancellation rate is 6% for both groups under the null. Then, use the standard formula for non-inferiority sample size for proportions: n per arm = ( (z_alpha + z_beta)^2 * (p1(1-p1) + p2(1-p2)) ) / (margin)^2, where p1 = p2 = 0.06, z_alpha = 1.645 (one-sided 0.05), z_beta = 0.84 (80% power), and margin = 0.005. Finally, compute the value and round up to the nearest integer.

Pro tip: In practice, always verify whether the margin is absolute or relative, and consider continuity correction or exact methods for small rates; also, be prepared to discuss how clustering or unequal allocation would affect the sample size.

1. Clarify the hypothesis and parameters

Confirm that the non-inferiority margin is 0.5 percentage points (absolute) and that the baseline rate is 6% for both arms under the null. Identify one-sided alpha = 0.05 and power = 80%.

2. Choose the appropriate formula

Use the sample size formula for non-inferiority of two proportions: n per arm = ( (z_alpha + z_beta)^2 * (p1(1-p1) + p2(1-p2)) ) / (delta)^2, where delta is the non-inferiority margin.

3. Plug in values and compute

Set p1 = p2 = 0.06, delta = 0.005, z_alpha = 1.645, z_beta = 0.84. Compute the numerator and denominator to find n per arm.

4. Round up and interpret

Round the calculated sample size up to the next whole number, as you cannot have a fraction of a participant. State the final per-arm sample size clearly.

Key Points to Mention

  • Non-inferiority margin is absolute (0.5 percentage points), not relative.
  • Use one-sided alpha = 0.05, corresponding to z = 1.645.
  • Power = 80% corresponds to z_beta = 0.84.
  • Baseline rate p = 0.06 for both arms under the null hypothesis.
  • Formula: n per arm = ( (z_alpha + z_beta)^2 * (p1(1-p1) + p2(1-p2)) ) / (margin)^2.
  • Round up to the nearest integer for the final sample size.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q3

You have one primary metric, two guardrail metrics, and one secondary metric. Propose and justify an error control strategy across these metric families and explain how you'd report adjusted confidence intervals.

A/B Testing & ExperimentationTechnical Trade-offs
Author's notes

I went with a gatekeeping structure: primary gets the full alpha budget at 0.05, guardrails get tested for non-inferiority at a shared 0.05 one-sided (no adjustment needed since they're non-inferiority claims), and the secondary only gets tested if the primary passes.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the metric hierarchy and the risks of multiple testing, then propose a gatekeeping strategy that controls family-wise error rate (FWER) across primary and secondary metrics while using a more lenient false discovery rate (FDR) or independent thresholds for guardrails. Justify the choice based on business impact and statistical power, and explain how to compute and report adjusted confidence intervals that reflect the chosen error control.

Pro tip: In practice, guardrail metrics are often monitored with one-sided tests and less stringent error control because the cost of a false alarm is lower than missing a regression; however, you must pre-register the strategy to avoid p-hacking accusations.

1. Clarify metric roles and testing risks

Define the primary metric as the key success measure, guardrails as safety checks, and secondary as supportive. Explain that testing multiple metrics inflates Type I error and requires a pre-specified error control plan.

2. Choose error control per family

For the primary and secondary metrics, use a gatekeeping procedure (e.g., fixed-sequence or Holm-Bonferroni) to control FWER at α=0.05. For guardrails, use a less stringent approach like Benjamini-Hochberg FDR or independent one-sided tests at α=0.10, justified by asymmetric costs.

3. Compute adjusted confidence intervals

For FWER-controlled metrics, adjust confidence levels using Bonferroni or Holm (e.g., 97.5% CI for primary if two metrics). For FDR-controlled guardrails, use Benjamini-Hochberg adjusted p-values and report CIs at the corresponding adjusted level (e.g., 95% CI but note FDR control).

4. Report and interpret results

Present unadjusted and adjusted CIs side-by-side, clearly labeling which metrics are confirmatory vs. exploratory. Emphasize that guardrail CIs are for monitoring, not strict hypothesis testing, and discuss any trade-offs in power.

Key Points to Mention

  • Family-wise error rate (FWER) vs. false discovery rate (FDR) and when to use each
  • Gatekeeping procedures (fixed-sequence, Holm-Bonferroni) for hierarchical testing
  • Bonferroni correction and its impact on confidence interval width
  • One-sided vs. two-sided tests for guardrail metrics
  • Pre-registration of the analysis plan to prevent p-hacking
  • Trade-off between Type I and Type II error rates and statistical power

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q4

Randomization is at the zone-day level with 300 orders per cluster and an ICC of 0.03. Compute the design effect and the adjusted per-arm sample size, then figure out how many zone-days per arm are needed.

A/B Testing & ExperimentationSystem Design
Author's notes

Design effect is 1 + (m-1)*ICC = 1 + 299*0.03 = about 9.97, so basically 10x.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

First, compute the design effect using the formula 1 + (m-1)*ICC, where m is the cluster size (300 orders per zone-day). Then, adjust the per-arm sample size by multiplying the standard sample size by the design effect. Finally, divide the adjusted sample size by the cluster size to get the number of zone-days per arm, rounding up to ensure sufficient power.

Pro tip: Always clarify whether the ICC is at the order or zone-day level; here it's zone-day, so the cluster size is the number of orders per zone-day. Also, remember that the design effect increases sample size, so the number of zone-days will be larger than if you ignored clustering.

1. Identify parameters

Extract the cluster size (m = 300 orders per zone-day) and the intra-cluster correlation (ICC = 0.03) from the problem statement.

2. Compute design effect

Use the formula DE = 1 + (m - 1) * ICC to calculate the design effect. Plug in m = 300 and ICC = 0.03 to get DE = 1 + 299 * 0.03 = 9.97.

3. Adjust per-arm sample size

Multiply the standard per-arm sample size (n) by the design effect to get the adjusted per-arm sample size: n_adj = n * DE. If n is not given, state that you need it, but typically in such problems n is provided or you can express the answer in terms of n.

4. Calculate zone-days per arm

Divide the adjusted per-arm sample size by the cluster size (m) to get the number of zone-days per arm: zone-days = n_adj / m. Round up to the nearest whole number.

5. Interpret and validate

Explain that the design effect inflates the sample size due to clustering, and the number of zone-days ensures adequate power. Mention that if n is not provided, you would need to compute it based on desired power and effect size.

Key Points to Mention

  • Design effect formula: DE = 1 + (m - 1) * ICC
  • Cluster size (m) is the number of orders per zone-day (300)
  • ICC is the intra-cluster correlation (0.03)
  • Adjusted sample size = standard sample size * design effect
  • Number of clusters (zone-days) per arm = adjusted sample size / cluster size
  • Rounding up to ensure sufficient sample size
  • If standard sample size is not given, you need to compute it using power analysis

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q5

With four equally spaced interim looks using O'Brien-Fleming alpha spending, explain how early stopping boundaries compare to the final look boundary, and describe how this affects minimum detectable effect and trial runtime in practice.

A/B Testing & ExperimentationTechnical Trade-offs
Author's notes

OBF is conservative early, so the first look has a critical value way out in the tail, something like z = 4 or higher, meaning you'd only stop early for an enormous effect.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by explaining that O'Brien-Fleming spending creates conservative early boundaries that are very stringent, while the final boundary is close to the unadjusted alpha. Then discuss how this design inflates the required sample size to maintain power, which increases the minimum detectable effect (MDE) for a fixed sample size or extends trial runtime to detect the same effect.

Pro tip: Emphasize that O'Brien-Fleming is often chosen for its conservative early stopping, which is ideal when early data is less reliable, but it comes at the cost of a larger maximum sample size. In practice, this means you need to plan for a longer trial or accept a larger MDE if you cannot increase the sample size.

1. Explain O'Brien-Fleming alpha spending

Describe how the alpha spending function allocates very little alpha to early looks and more to later looks, resulting in stringent early boundaries that are hard to cross.

2. Compare early vs. final boundaries

Contrast the early stopping boundaries (e.g., very small p-value thresholds) with the final look boundary, which is close to the nominal alpha (e.g., 0.05) but slightly adjusted for multiple looks.

3. Discuss impact on minimum detectable effect (MDE)

Explain that because early boundaries are stringent, the trial needs a larger sample size to maintain power, which increases the MDE for a fixed sample size or requires a larger sample to detect the same effect.

4. Describe effect on trial runtime

Highlight that to achieve the same power and MDE, the trial must run longer to accumulate the required sample size, unless the effect is large enough to stop early.

5. Summarize practical trade-offs

Conclude that O'Brien-Fleming balances early stopping conservatism with minimal alpha penalty at the final look, but requires careful planning of sample size and runtime to avoid underpowered studies.

Key Points to Mention

  • O'Brien-Fleming spending function: alpha spent proportional to sqrt(t) or similar, with very small alpha at early looks.
  • Early boundaries are extremely conservative (e.g., p < 0.0001), while final boundary is close to 0.05.
  • Multiple testing adjustment inflates maximum sample size by a factor (e.g., 1.02-1.1) to maintain power.
  • Minimum detectable effect increases if sample size is fixed, or trial duration increases to detect the same effect.
  • Early stopping is only likely if the treatment effect is very large, which is rare.
  • Practical implication: plan for a longer trial or larger sample size when using O'Brien-Fleming.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.